Detroit Police Wrongfully Arrested Black Man Due To Faulty FRT

In January 2020, Robert Julian-Borchak Williams, a 42-year-old Black man from Farmington Hills, Michigan, was wrongfully arrested by Detroit police for allegedly stealing $3,800 worth of watches from a Shinola store in October 2018. The arrest occurred after Michigan State Police used facial recognition technology from DataWorks Plus to analyze grainy surveillance footage from the store, which incorrectly identified Williams' old driver's license photo as a match for the suspect. The facial recognition system incorporated algorithms from NEC and Rank One Computing, both of which had been found in federal studies to falsely identify African-American faces 10 to 100 times more than Caucasian faces. Williams was arrested at his home in front of his wife and two young daughters, held for 30 hours in detention, and released on $1,000 bond. During interrogation, detectives showed Williams the surveillance photos, and he pointed out the obvious differences between himself and the actual suspect. One detective acknowledged 'the computer got it wrong.' The case was dismissed by prosecutors two weeks later due to insufficient evidence. Williams' case is considered the first documented instance of a wrongful arrest in the United States based on faulty facial recognition technology. The incident also affected two other Black men, Michael Oliver and Nijeer Parks, who were similarly wrongfully arrested due to facial recognition errors by the same detective system.